The Journal of Pain, Vol 16, No 12 (December), 2015: pp 1221-1232

Available online at www.jpain.org and www.sciencedirect.com

Critical Review
Inhaled Cannabis for Chronic Neuropathic Pain: A Meta-analysis
of Individual Patient Data
Michael H. Andreae,* George M. Carter,y Naum Shaparin,* Kathryn Suslov,z
Ronald J. Ellis,x Mark A. Ware,{ Donald I. Abrams,k Hannah Prasad,** Barth Wilsey,**
Debbie Indyk,z Matthew Johnson,yy and Henry S. Sacksz
*Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, New York.
y
Foundation for Integrative AIDS Research, Brooklyn, New York.
z
Icahn School of Medicine at Mount Sinai, New York, New York.
x
Department of Neurosciences, University of California, San Diego, California.
{
Department of Anesthesia and Family Medicine, McGill University, Montre al, Que
bec, Canada.
k
AIDS Program, San Francisco General Hospital, University of California, San Francisco, California.
**Department of Physical Medicine and Rehabilitation, VA Northern California and University of California Davis
Medical Center, Sacramento, California.
yy
Teachers College, Columbia University, New York, New York.

Abstract: Chronic neuropathic pain, the most frequent condition affecting the peripheral nervous sys-
tem, remains underdiagnosed and difficult to treat. Inhaled cannabis may alleviate chronic neuropathic
pain. Our objective was to synthesize the evidence on the use of inhaled cannabis for chronic neuro-
pathic pain. We performed a systematic review and a meta-analysis of individual patient data. We regis-
tered our protocol with PROSPERO CRD42011001182. We searched in Cochrane Central, PubMed,
EMBASE, and AMED. We considered all randomized controlled trials investigating chronic painful neu-
ropathy and comparing inhaled cannabis with placebo. We pooled treatment effects following a hier-
archical random-effects Bayesian responder model for the population-averaged subject-specific effect.
Our evidence synthesis of individual patient data from 178 participants with 405 observed responses in
5 randomized controlled trials following patients for days to weeks provides evidence that inhaled
cannabis results in short-term reductions in chronic neuropathic pain for 1 in every 5 to 6 patients
treated (number needed to treat = 5.6 with a Bayesian 95% credible interval ranging between 3.4
and 14). Our inferences were insensitive to model assumptions, priors, and parameter choices. We
caution that the small number of studies and participants, the short follow-up, shortcomings in alloca-
tion concealment, and considerable attrition limit the conclusions that can be drawn from the review.
The Bayes factor is 332, corresponding to a posterior probability of effect of 99.7%.
Perspective: This novel Bayesian meta-analysis of individual patient data from 5 randomized trials
suggests that inhaled cannabis may provide short-term relief for 1 in 5 to 6 patients with neuropathic
pain. Pragmatic trials are needed to evaluate the long-term benefits and risks of this treatment.
ª 2015 by the American Pain Society
Key words: Cannabis, chronic pain, neuropathy, painful, polyneuropathy, meta-analysis, meta-analysis
of individual patient data, Bayesian analysis, human immunodeficiency virus.

This work was supported by the National Center for Advancing Transla-
tional Sciences (NCATS), a component of the National Institutes of Health
(NIH), through CTSA grant numbers UL1TR000086, TL1RR000087, and
KL2TR000088), the Center for Drug Evaluation and Research (CDER)
through grant number R01-AT005824 and in part by Grant 5R01AT5824
from the National Center for Complementary and Alternative Medicine
(NCCAM). Supported by the University of California Center for Medicinal
Cannabis Research and NIH Grant 5-MO1-RR00083. Its contents are solely
the responsibility of the authors and do not necessary represent the offi-
cial view of the NCCAM, NCRR or NIH. The authors have no conflicts of
interest.
Supplementary data accompanying this article are available online at
www.jpain.org and www.sciencedirect.com.
Address reprint requests to Michael H. Andreae, MD, Department of
Anesthesiology, Montefiore Medical Center, Albert Einstein College of
Medicine, 111 E 210th St, Bronx, NY 10467. E-mail: mhandreae@gmail.
com
1526-5900/$36.00
ª 2015 by the American Pain Society
http://dx.doi.org/10.1016/j.jpain.2015.07.009

1221
1222 The Journal of Pain Inhaled Cannabis for Chronic Neuropathy
Methods
A
bout 1 in 40 adults in the general population has
chronic neuropathic pain, making it the most
We registered our protocol with PROSPERO.5 We identi-
frequent condition affecting the peripheral ner-
fied studies by a combination of electronic and manual
vous system.52 Chronic neuropathic pain presents a het-
searches (Fig 1) (Supplementary Appendix 1). We followed
erogeneous burden with a large prevalence12 in certain
the recommendations of the QUORUM and PRISMA state-
susceptible subpopulations, for example in people living
ments,59 including the PRISMA checklist (Supplementary
with human immunodeficiency virus (HIV).30 HIV-related
Appendix 2). We searched in Cochrane Central, PubMed,
distal sensory neuropathy affects every third patient.30
EMBASE, and AMED without any language restriction
Chronic neuropathic pain may result from diverse insults,
with a combination of free text and controlled vocabulary,
including diabetes, HIV, trauma, and certain medica-
using the highly sensitive search strategy.44 We conducted
tions.86,87 Chronic neuropathic pain remains
a hand search in the conference abstracts from the Confer-
underdiagnosed and difficult to treat.33 Regardless of
ence on Retroviruses and Opportunistic Infections 2011,
the cause, chronic neuropathic pain persists despite at-
the International AIDS Conference, and the World
tempts at management with opioids, nonsteroidal anti-
Congress of Pain 2010 and reference lists.
inflammatory drugs, anticonvulsants (gabapentin),
We considered RCTs investigating chronic painful neu-
anti-inflammatory agents, antidepressants, and comple-
ropathy. We included diabetic, traumatic, and HIV-
mentary medicines.33
related causes. We excluded multiple sclerosis, a central
A recent systematic review concluded that cannabis is
rather than a peripheral pain condition. The nature of
effective in selected neurological disorders, including
the intervention likely interfered with effective partici-
multiple sclerosis, but did not address chronic neuro-
pant blinding,3 which was therefore not required for
pathic pain.50 Considering the recent wave of cannabis
study inclusion. We only included studies comparing
legalization,76 continued legal wrangling,65 its wide-
inhaled Cannabis sativa with placebo, because inhaled
spread medicinal and recreational use,80,88 and
whole-herb cannabis differs significantly in composition,
additional randomized controlled trials (RCTs)
bioavailability, and pharmacodynamics from synthetic
published on cannabis recently, we performed a meta-
cannabinoids.70
analysis to investigate if inhaled cannabis alleviates
Three review authors (M.H.A., G.C., K.S.) screened the
chronic neuropathic pain.8,81,84 Previous (systematic)
citations using explicit criteria for study exclusion. Using
reviews did not investigate inhaled cannabis for
a standard data collection form, 2 authors (M.H.A. and
chronic neuropathic pain or were unable to synthesize
G.C.) extracted the data independently, reconciling any
all available data, did not include recently published
differences by consensus. Study authors provided indi-
RCTs, and varied considerably in their inclusion criteria,
vidual patient data.2,31,81,84,85 We recorded details of
study selection, and data synthesis, leading to
trial design, conflict of interests, sponsors, participant
conflicting and outdated conclusions.13,19,23,47,50,54-
56,63,67,76,79,89
As cannabis should undergo the same
evidence-based review39 as other potent prescription
medications,83 an update is needed.81,84
In cooperation with all primary study authors, we per-
formed an individual patient data Bayesian35 meta-
analysis of RCTs71 (Supplementary Box 1). Although
classic meta-analysis pools aggregate data extracted
from published study reports, meta-analysis of individual
patient data synthesizes the original data of the individ-
ual participants obtained from the primary study au-
thors.75 This often gives meta-analysis of individual
patient data more power.75 We selected Bayesian evi-
dence synthesis for the analysis, anticipating that incom-
plete outcome reporting, varied endpoints, limited
availability of aggregate or individual patient data, and
diversity of study designs with varied statistical analysis
approaches would pose a formidable challenge to the
classic (frequentist) methods of meta-analysis.74 Classic
meta-analysis may also underestimate the between-
study variability for small numbers of trials,24,73 leading
to inaccurate inferences; Bayesian methods provide
more robust estimates of between-study variance.
Figure 1. The QUORUM flow chart details our search in a di-
agram. We selected 165 articles for full review from 1738 hits
Objectives in multiple electronic databases; 5 RCTs met the inclusion
criteria (Abrams 2007,2 Ellis 2009,31 Ware 2010,81 Wilsey
We performed a Bayesian responder meta-analysis of
2008,85 Wilsey 201384). Excluded studies are double counted
individual patient data to study whether inhaled if they met more than 1 exclusion criterion, eg disease and
cannabis provides relief for chronic neuropathic pain. mode of administration.
Andreae et al The Journal of Pain 1223
characteristics, interventions and outcome measures,
inclusion and exclusion criteria, comorbidity and HIV
status, cannabis provenance, dose, and mode of
administration. We extracted data on attrition and on
adverse effects.
We compared the proportion of patients experiencing
more than 30% clinical improvement in chronic neuro-
pathic pain assessed with a continuous patient-reported
instrument (eg, the visual analog scale) at baseline and af-
ter treatment with inhaled cannabis. In essence, we
dichotomized the outcome in a responder analysis,
emerging as the preferred method for pain outcomes
research.28,32 We chose this patient-centered concept of
minimally clinically important difference,58 because
chronic neuropathic pain, our primary outcome, is patient
reported and may have a skewed distribution, with no
more than 40 to 60% of patients obtaining even partial
relief of their pain.27 A statistically significant change in
the population mean of a continuous pain outcome
may not correspond to a clinically meaningful improve-
ment for many individual patients.60 In other words, large
studies may detect population differences too small for
individual patients to appreciate. However, responder
analysis converts continuous pain outcomes to dichoto-
mous responder data allowing a more meaningful com-
parison between interventions.61,72 By convention, we
classified participants as ‘‘responders’’ if their reduction
Figure 2. This summary of bias graph shows that the studies
in continuous spontaneous pain outcome (eg, VAS included were mostly of good quality in the domains of
score) was larger than 30% after treatment.28,32 sequence generation, concealed allocation, incomplete
Two authors (G.C. and M.H.A.) independently assessed outcome data, and selective reporting, and with regard to con-
the risk of bias of the studies included according to the flict of interest. However, the nature of the intervention likely
interfered with effective blinding, possibly resulting in a high
Cochrane Collaboration44 on the basis of a checklist of risk of performance bias in all studies and detection bias due
design components and they contacted authors for to a lack of blinding of outcome observers. (Abrams 2007,2 Ellis
missing information. We summarized this in a risk of 2009,31 Ware 2010,81 Wilsey 2008,85 Wilsey 201384).
bias graph (Fig 2) and provide detailed information in
Supplementary Table 1. This comprised randomization, detailed in the statistical supplement (Supplementary
allocation concealment, observer blinding, intention- Appendix 3). We pooled the treatment effects following
to-treat analysis, selective reporting, and conflict of in- a hierarchical random-effects Bayesian responder model.
terests. We achieved consensus by informal discussion. Kruschke51 provided an accessible introduction to
With inhaled cannabis interventions, blinding of pa- Bayesian methods in health sciences. Ashby6 recently
tients and providers can be difficult and hence they offered a chronological outline of applications in medi-
received less weight in the evaluation of performance cine, and Spiegelhalter et al74 compiled the first concise
bias but not with regard to detection bias. overview. Gelman et al35 described Bayesian hierarchical
Our results are based on individual patient data obtained modeling approaches more formally. Supplementary
from primary authors who helped resolve data inconsis- boxes explain the basic concepts of Bayesian inference
tencies when evident. We estimated the content and the (Supplementary Boxes 1–3). The prior for the between-
dose administered following published methods10,57 in study variability (Cauchy) and the pooled effect estimate
cooperation with the primary study authors. (normal distribution) were centered at zero with a stan-
We compared the reported primary outcome with the dard deviation of 100. We preferred the Cauchy distribu-
planned primary outcome in the study protocols to assess tion over the closely related t-distribution, because the
reporting bias. We explored undue sponsor influence.44 Cauchy is more robust in accommodating outliers34,35;
We considered an examination of publication bias using these priors for our meta-analysis were uninformative
graphical and statistical tests.29 We investigated study and served to ensure computational convergence of
heterogeneity using a c2 test and calculation of an I2 the Markov chain Monte Carlo algorithm. Our priors
analog Bayesian statistic.44 were subsequently subjected to sensitivity analysis. Infer-
ence was implemented using a Gibbs sampling scheme to
generate a computer simulation of a Monte Carlo sample
Data Synthesis, Statistical Modeling, and
from the posterior distribution in OpenBugs.53 Our
Sensitivity Analysis OpenBugs program code is provided in Supplementary
We performed full Bayesian probability modeling20 of Appendix 4. We have uploaded details on Monte Carlo
the population-averaged subject-specific effect90 as Markov chain convergence, including graphs
1224 The Journal of Pain
demonstrating mixing, as supplementary material
(Supplementary Figs 1 and 2). Differences in the design
and quality of the studies were the focus of a sensitivity
analysis. We tested the sensitivity of our results for our
Bayesian model and its assumptions. We investigated
our choice of prior and model parameters and reana-
lyzed the individual patient responder data 1) in a fre-
quentist random-effects meta-analysis and 2)
controlling for cannabis dose as an explanatory variable
of the between-study variability in a meta-regression
(methods and data not shown but available on request).
Reporting
We estimated the number needed to treat (NNT) and
calculated the Bayes factor,36 compared with the classic
P values in Supplementary Box 3. We provided forest
plots for the individual trials broken down by dose for
the period level data (Fig 3). The reported pooled
Bayesian estimate is the population-averaged subject-
specific odds ratio comparing inhaled cannabis versus
placebo for chronic painful neuropathy and their 95%
Bayesian credible intervals (CRI95%), displayed as the
standard diamond used for synthesized effects.43
Differences From the Initial Protocol
In our initial Prospero protocol registration, we consid-
ered including all types of studies, populations, and
cannabis interventions. We intended to do a network
analysis in 1 coherent Bayesian model. We found pub-
lished aggregate data insufficient for evidence synthesis
and therefore we decided to attempt an individual pa-
tient data meta-analysis, but limiting ourselves to only
Inhaled Cannabis for Chronic Neuropathy
RCTs investigating inhaled cannabis, and updated the
protocol accordingly.
Results
Our search (Fig 1) was completed in April 2014 and
yielded 1738 references (1236 in MEDLINE, 359 in EM-
BASE, 123 in Cochrane Central, and 65 in AMED) matching
the predefined search parameters. We excluded 118 du-
plicates and 1573 references for which we could clearly
discern from the title or abstract that they were not ran-
domized trials or did not investigate cannabis for a pain-
ful condition. Our hand search yielded no additional
references. Except for the 5 studies included,2,31,81,84,85
the remaining 163 publications studied different modes
of cannabis administration or included participants with
other painful conditions. No study investigated
outcomes beyond 2 weeks. The characteristics of the 5
RCTs meeting our inclusion criteria are summarized in
Table 1 and detailed characteristics are presented in
Table 2. Important studies that were excluded are listed
in Supplementary Table 2 with reasons for their exclusion.
Descriptive Characteristics of the Studies
Included and the Participants
One hundred and seventy-eight middle-aged partici-
pants (approximately equal numbers of men and
women) with painful neuropathy of at least 3 months
duration (pain scores at least about 3/10) were enrolled
in 5 RCTs executed across North America. Two trials re-
cruited only HIV-positive individuals with HIV-related
chronic painful neuropathy1,2,31; sexual orientation and

Study Dose Placebo Treat Est. OR (CI)

Abrams 07 96 6/25 13/25 3.43 (1.00,11.8)

Ellis 09 96 5/28 13/28 5.00 (1.10,22.9)

Ware 10 2.5 3/22 4/21 1.50 (0.25,8.98)

Ware 10 6.3 5/22 3.00 (0.31,28.8)
Ware 10 9.4 7/21 5.00 (0.58,42.8)

Wilsey 08 19 18/33 24/36 2.67 (0.71,10.1)

Wilsey 08 34 22/33 3.50 (0.73,16.8)

Wilsey 13 9 11/38 17/37 2.50 (0.78,7.97)

Wilsey 13 18 18/36 3.67 (1.02,13.1)

Bayesian 3.22 (1.59,7.24)

0.2 0.5 1.0 2.0 5.0 10.0 20.0 40.0

Odds Ratio

Figure 3. The forest plot displays odds ratios (with the 95% CRI indicated by horizontal bars on the log scale) to indicate their contri-
bution to the Bayesian pooled effect estimate shown below as a diamond with the Bayesian 95% CRI. The table on the left lists the raw
responder data at the study level. For Ware 2010,81 Wilsey 200885 and Wilsey 2013,84 the responder data are broken down by dose,
listing the number of observed responses for each crossover period and the corresponding cannabis dose. The increased effect with
increased cannabis content (evident in the period level data of Ware 2010,81 Wilsey 2008,85 and Wilsey 201384) is additional evidence
in support of the effect of cannabis on chronic painful neuropathy.
Andreae et al
Table 1. Summary of the 5 RCTs on Inhaled Cannabis for Chronic Neuropathic Pain
CHARACTERISTIC ABRAMS 2007
Neuropathy HIV-DSPN
Participants 50
Allocation <-------------------------------------------------------------- Randomized --------------------------------------------------------------->
Intervention <-------------------------------------------------- Inhaled cannabis versus placebo -------------------------------------------------->
Outcome VAS
Follow-up <----------------------- 5 days --------------------->
Design Parallel
Statistics Mann-Whitney
Abbreviations: HIV-DSPN, HIV-related distal sensory polyneuropathy; DDS, Descriptor Differential Scale; NRS, Numerical Rating Scale.
NOTE. The cause of chronic neuropathic pain varied (including traumatic, central, diabetic, and HIV-related). Study authors used several patient-reported pain outcome
instruments.
transgender data were not reported. Three trials
recruited patients with neuropathy secondary to
trauma,81 spinal cord injury, diabetes mellitus, and com-
plex regional pain syndrome.84,85 Psychiatric disease,
substance abuse, and significant cardiopulmonary
disease were explicit exclusion criteria. Although
previous cannabis experience was a prerequisite for
inclusion for some studies,1,2,84,85 current use was an
exclusion criterion in all. Prescribed opioid use was not
specified among the inclusion or exclusion criteria.
Characteristics of Interventions and
Comparators
All studies investigated inhaled cannabis. The 5 studies
used different doses, estimated as detailed in the
Supplementary Table 3. All 5 studies used whole Cannabis
plant provided by the US National Institute of Drug Abuse
(NIDA). Cannabis was administered as prerolled cigarettes
in 3 studies,1,2,31,85 through a Volcano vaporizer in 1
study,84 and as gelatin capsules smoked through a pipe
at home in 1 study.81 All 5 studies used identical-looking
placebo as comparator. Concomitant nonstudy analgesics
were permitted and continued in both arms.
Clinical Outcomes and Adverse Effects
All 5 RCTs reported continuous patient-reported spon-
taneous pain intensity scales as primary outcomes. We
report the study level observed odds ratio (with 95%
CRI indicted by vertical bars on the log scale) as a measure
of their contribution to the Bayesian pooled effect esti-
mate (shown as a diamond with 95% CRIs below) (Fig
3). The breakdown of responder data by dose suggested
an increased effect with increased cannabis content.
Withdrawals due to adverse effects were rare. One
case of serious adverse effects leading to withdrawal
occurred in the placebo group (a case of psychosis) and
2 others in active treatment groups (hypertension and
increased pain). Subjective side effects included anxiety,
disorientation, difficulty concentrating, headache, dry
eyes, burning sensation, dizziness, and numbness and
were reported as being mild. Wilsey et al85 reported
short-term declines in attention, psychomotor perfor-
mance, and learning and memory in the highest dose
(7% tetrahydrocannabinol) group. Memory impairment
was also seen in the placebo group and at lower doses,
albeit at lower levels. Statistically significant physiolog-
The Journal of Pain 1225
ELLIS 2008 WARE 2010 WILSEY 2008 WILSEY 2013
HIV-DSPN Posttraumatic Sensory Mixed
34 23 38 39
DDS NRS VAS VAS
2 weeks <------------------ 5–6 h ---------------->
<------------------------------------------------ Crossover ------------------------------------------------>
Wilcoxon rank sum <------------- General and linear mixed models ------------->
ical changes (such as increases in heart rate) were
observed in one study31 but not in another study.81
Reports of euphoria or ‘‘high’’ were rare.81 Psychoactive
effects (such as feeling ‘‘high’’) were statistically signifi-
cantly associated with treatment allocation in 2
studies2,31 and increased in frequency with increasing
dose81,85; they were mostly mild. The studies included
followed patients only for days to weeks and hence did
not report long-term adverse effects.
Study Design
All 5 studies were randomized, placebo-controlled,
and double-blind; 4 used a crossover design31,81,84,85
and 1 study used a parallel design.1 Duration of follow-
up varied from hours84,85 to days2,31 or weeks.81
Risk of Bias in the Studies Included
We characterized the risk of bias of the studies (Fig 2
and Supplementary Table 1). Randomization and alloca-
tion concealment were well described and suggested a
low risk of bias. Ineffective participant blinding might
have possibly resulted in performance bias in all studies;
placebo effects are likely, when participants guessed
their allocation, possibly leading them to overestimate
the effect of inhaled cannabis on pain. Blinding of
outcome observer was well described in 1 study,84 and
the use of patient diaries as an outcome instrument led
us to estimate the risk of detection bias as unclear in
the remaining studies. Incomplete outcome data were
well described in all studies and are detailed in Table 2.
Withdrawals potentially related to treatment effects
led to a high risk of bias in 1 study81 but did not seem
to be associated with group allocation in all
others.2,31,84,85 All the trials included reported their
primary outcome as specified in the protocol. We
investigated publication bias in a funnel plot proposed
by Egger et al,29 because with fewer studies than 10
studies, the power of the tests is insufficient to distin-
guish chance from real asymmetry.44 Studies received
only public funding; all authors provided detailed con-
flicts of interest statements.
Evidence Synthesis of Effects
Based on data from 178 patients with a total of 405
observed responses, we estimated the odds ratio for
more than 30% reduction in pain scores in response to
1226 The Journal of Pain Inhaled Cannabis for Chronic Neuropathy
Table 2.Detailed Characteristics of the 5 RCTs Investigating Smoked or Inhaled Cannabis for
Painful Neuropathy
STUDY ID YEAR JOURNAL PUBMED ID TRIAL REGISTRY ID
2
Abrams 2007 2007 Neurology 17296917 NCT00046722
Population 55 HIV-positive adults with symptomatic HIV-DSPN and at least 30/100 VAS, on stable pain regimen for 8 weeks before
enrolment, with previous experience of smoking cannabis randomized in 2 groups of size 27/28. Our Bayesian analysis
is based on 50 participants with 1 observation per patient, as provided by the primary study authors.
Age (experimental, control): 50 y (SD 6 6 y), 47 y (SD 6 7 y)
Gender (male/female/other): experimental 22/5/0, control 26/2/0
Intervention Experimental: patient smoked 1 cigarette 3 times per d as tolerated
Prerolled, whole-herb Cannabis cigarettes were provided by NIDA and contained 3.56% D-9-THC.
Control: identical prerolled cigarettes with the active ingredient extracted.
Dose estimate: 32 mg THC per session; 96 mg THC per d
Primary outcome Daily pain diary recording the VAS at 8 AM for average pain during the previous 24 h
Study methods Randomized, double-blind (patient, outcome assessor), parallel design, placebo-controlled, single-center (university) clinical
trial in San Francisco, California, starting in 2003
Notes Also published as an abstract at the 2nd Annual Meeting of the International Association for Cannabis as Medicine, 2005
Secondary outcomes: acute analgesic effects; long thermal stimulation
Anti-hyperalgesic effects: heat-capsaicin model; profile of mood states
Ellis 200931 2008 Neuropsychopharmacology 18688212 NCT00255580
Population 34 HIV-positive adults with symptomatic HIV-DSPN and pain score >5/20 on DDS, most participants were previously
exposed to potentially neurotoxic deoxy-nucleoside reverse transcriptase inhibitors; 16 started control/experimental, 18
started experimental/control. 28 participants with a total of 56 observed responses were included in the Bayesian analysis
Age (all): 49.1 y (SD 6 6.9 y)
Gender (male/female/other): 33/1/0
Intervention Experimental: patient smoked cannabis, titrating dose up or down to obtain effective pain control/tolerable adverse effects,
starting at 4% and ranging between 1 and 8% D-9-THC concentration by weight. Prerolled, whole-herb Cannabis
cigarettes were provided by NIDA
Control: identical prerolled cigarettes with the active ingredient extracted
Dose estimate: average 96 mg THC per d
Primary outcome Crossover difference in change of DDS 0–20 scale ‘‘a ratio scale containing 24 words describing pain intensity and
unpleasantness’’ comparing baseline with after treatment
Study methods Randomized, double-blind (patient, outcome assessor), crossover design, placebo-controlled, single-center (university)
clinical trial at the University of California, San Diego, California, in 2006
Notes Secondary outcomes: McGill questionnaire, VAS, Sickness Impact Profile, Brief Symptom Inventory, UKU side effect rating,
Highness/Sedation Scale, HIV load
Ware 201081 2010 Canadian Medical 20805210 ISRCTN68314063
Association Journal
Population 23 adults with non-HIV neuropathy pain of at least 3 mo duration caused by trauma or surgery defined by pain intensity
score greater than 40/100 VAS, on a stable analgesic regimen, not having smoked cannabis in the preceding year. 23
participants with a total of 86 observed responses were included in the Bayesian analysis
Age (all): 45.4 y (SD 6 12.3 y)
Gender (male/female/other): 11/12/0
Intervention Experimental: NIDA and Prairie plant systems prepared 3 different potencies of THC (2.5%, 6%, 9.4%) from whole herb
in gelatin capsules inhaled through a pipe
Control: ethanolic extraction was used to prepare the placebo
Dose estimate: 0, 1.625, 3.9, and 5.85 mg/d (average) THC per period
Primary outcome Average daily pain intensity on the 11-item NRS averaged over 5 treatment d (least pain value, average pain value, and
worst pain value) during 4 consecutive crossover periods of 14 d each (5 treatment d and 9 washout d afterwards)
Study methods Randomized, double-blind (patient, outcome assessor), 4-period crossover Latin square design, placebo-controlled,
single-center (university) clinical trial in McGill University, Montreal, Canada, starting in 2003
Notes The linear model did not consider interparticipant effects
Secondary outcomes: pain quality, McGill questionnaire, sleep (Leeds Sleep Evaluation Questionnaire); mood effects,
short-form Profile of Mood States; quality of life: EQ-5D health outcomes
Wilsey 200885 2008 The Journal of Pain 18403272 NCT00254761
Population 38 adults with non-HIV neuropathy (complex regional pain syndrome (type I), spinal cord injury, peripheral neuropathy, or
nerve injury) with previous cannabis experience and a VAS >30/100. 38 participants with 102 observed responses were
included in the Bayesian analysis
Age (all): 46 y (range 21–71 y)
Gender (male/female/other): 20/18/0
Andreae et al The Journal of Pain 1227
Table 2. Continued
STUDY ID YEAR JOURNAL PUBMED ID TRIAL REGISTRY ID
Intervention Experimental participants inhaled a total of 9 standardized cued puffs. Cannabis was harvested from whole plant and
rolled into cigarettes at the University of Mississippi under supervision of NIDA ranging in strength from 0% to 3.5–7%.
Control: placebo cigarettes were made from whole plant with extraction of Cannabis
Dose estimate: 0 placebo, 19.25 (low dose, range 7–30.45), 34.3 (high dose, range 18.9–60.9) mg THC/d (session)
Primary outcome VAS measuring spontaneous pain relief; time effects were studied with a linear model
Study methods Randomized, double-blind (patient, outcome assessor), parallel design, placebo-controlled, single-center (university) clinical
trial at University of California Davis Medical Center, Sacramento, California, started in November 2003
Notes Secondary outcomes: pain unpleasantness (VAS), heat pain threshold, Neuropathic Pain Scale, neurocognitive assessment,
and plasma Cannabis concentration
Wilsey 201384 2013 The Journal of Pain 23237736 NCT01037088
Population 39 adults with non-HIV neuropathy due to reflex sympathetic dystrophy, peripheral neuropathy, postherpetic neuralgia,
poststroke pain, multiple sclerosis, or spinal cord injury with previous cannabis exposure (16 current, 23 ex-users) and a
VAS pain intensity greater than 30/100. 39 participants with 111 observed responses were included in the Bayesian
analysis
Age (all): 50 y (SD 6 11 y)
Gender (male/female/other): 28/11/0
Intervention Experimental: participants used a volcano vaporizer under the flexible-dose design of Wilsey 2008. The minimum and
maximum cumulative doses for each visit were 8 and 12 puffs. Cannabis was harvested at the University of Mississippi
under the supervision of NIDA
Control: placebo was made from whole plant with removal of cannabinoids
Dose estimate: maximum of 0, 10.32, 28 mg THC/d (session), presuming they were administered the entire 800 mg dose
Primary outcome VAS before and after consuming vaporized cannabis
Study methods Randomized, double-blind (patient, outcome assessor), crossover design, placebo-controlled, single-center (university)
clinical trial at the University of California, Davis, California, started December 2009
Notes Secondary outcomes: Patient Global Impression of Change; Neuropathic Pain Scale; WAIS-III, Hopkins Verbal Learning Test
(revised), Grooved Pegboard Test

Abbreviations: HIV-DSPN, HIV-related distal sensory polyneuropathy; SD, standard deviation; DDS, Descriptor Differential Scale; NRS, Numerical Rating Scale.
NOTE. Two trials recruited patients with HIV-DSPN, 3 included participants with neuropathies due to other causes.

inhaled cannabis versus placebo for chronic painful neu-
ropathy as 3.2 with a Bayesian CRI (subsequently de-
noted with the subscript CRI95%) [1.59, 7.24]CRI 95%, and
the NNT as 5.55 [3.35, 13.7]CRI 95%. We estimated the pos-
terior probability of the effect of Cannabis for chronic
painful neuropathy to be 99.7% and the Bayes factor
as 332 (Fig 3). The Bayesian analog I2 statistic was 0.
The posterior probability that the between-study vari-
ability in effects was greater than what would be ex-
pected by chance is .45. Effects seemed to increase with
tetrahydrocannabinol (THC) content supporting the ef-
fect of cannabis for chronic painful neuropathy as seen
in the forest plot (Fig 3). Specifically, the increased effect
with increased cannabis content (evident in the period
level data of Ware 2010,81 Wilsey 200885 and Wilsey
201384) is additional evidence consistent with the effect
of cannabis for chronic painful neuropathy. However, a
meta-regression of cannabis dose (data not shown but
available on request) did not change our estimates or in-
ferences. The aggregate and individual data on adverse
effects were too sparse to be pooled. Model convergence
is documented in the supplementary material
(Supplementary Figs 1 and 2).
Sensitivity Analysis
When we performed a sensitivity analysis (available on
request) with regard to differences in the quality of
studies, we found effect estimates and credible intervals
to be robust regarding the inclusion or exclusion of any
single study. Our inferences were rather insensitive to
priors (between-study variance) in our Bayesian model
(Supplementary Box 2). Reanalyzing the data in a fre-
quentist random-effects meta-analysis did not change
the results.
Discussion
Our evidence synthesis of individual patient data from
178 participants with 405 observations in 5 RCTs with a
follow-up ranging from days to weeks (Fig 3) provides
evidence that inhaled cannabis results in short-term re-
ductions in chronic neuropathic pain for 1 in every 5 to
6 patients treated (NNT 5.6 with a Bayesian 95% CRI
ranging between 3.4 and 14); based on the Initiative
on Methods, Measurement, and Pain Assessment in Clin-
ical Trials (IMMPACT) definition of at least moderate
benefit,28 inhaled cannabis improved pain by an odds ra-
tios of 3.2 (Bayesian 95% CRI of [1.6, 7.2]CRI 95% (Fig 3).
The Bayes factor was 332, corresponding to a posterior
probability of effect of 99.7%.
We infer that this effect applies equally across chronic
painful neuropathies of different causes (eg, diabetic
and traumatic chronic painful neuropathy or HIV-
related distal sensory neuropathy). The effects are
remarkably homogeneous across studies (Bayesian I2
analog = 0%) (Table 1). Dose dependency further sup-
ports the notion of a cannabis effect on neuropathy
(Fig 3 and Supplementary Table 3). Our results (NNT 5.6
[3.4, 14]CRI 95%) suggest that inhaled cannabis may be
about as potent as gabapentin (Cochrane Review
1228 The Journal of Pain
update: NNT 5.9 [4.6, 8.3]CI 95% for diabetic neuropathy,
Moore 201462). The NNT of inhaled cannabis could
potentially rival currently available therapeutics for
chronic neuropathic pain,82 whose NNT typically range
well above 8, if there is any evidence at all.15,21,25
However, we caution that our findings await
confirmation in long-term pragmatic community-based
trials. Our findings are remarkable considering the
dearth of effective treatment options for chronic painful
neuropathies or chronic pain in general.49
Our Review Enhances the Existing
Literature on Treatment for Chronic
Neuropathic Pain
Our evidence synthesis contradicts, updates, or com-
plements the finding of several older and more recent re-
views on cannabis by providing a meta-analysis for
chronic neuropathic pain,13,14,85 by updating
evidence, 13-16
or by broadening the scope. We were
able to include recent RCTs, not published or accessible
to the previous reviews by Campbell, Phillips, Iskedjian,
Lutge, or Lynch.19,47,54,55,67 Compared with previous
studies,67 our meta-analysis of individual patient data
and the inclusion of additional and recent clinical trials,
which augmented the power to detect an effect, if it ex-
isted, and amplified the confidence in the pooled effect
estimate (NNT = 5.6) by shrinking the 95% CRI,2,3,12 our
posterior probability of the short-term effects of inhaled
cannabis is now very high (99.7%). Our analysis comple-
ments the recent evidence synthesis of cannabis for
certain other neurological conditions by the American
Academy of Neurology, which did not investigate
cannabis for chronic neuropathic pain,49 and supports
another narrative review (published after submission of
this manuscript) with a meta-analysis of individual pa-
tient data,45 which concluded that ‘‘Use of marijuana
for chronic pain, neuropathic pain, and spasticity due to
multiple sclerosis is supported by high-quality evidence.’’
Strength
Meta-analysis of Individual Patient Data
Increased the Power of our Meta-analysis
We performed an individual patient meta-analysis. Un-
like conventional meta-analysis based on published
aggregate data, meta-analysis of individual patient
data synthesizes the individual participants’ original
data obtained from the studies’ principal investigators.43
Meta-analysis of individual patient data is arguably the
gold-standard of evidence synthesis,71,75 not just
because it allows for detailed data checking but
because meta-analysis is often not feasible using only
summary data. Synthesizing the diversity of reported
outcomes of studies on inhaled cannabis for chronic
painful neuropathy was a significant challenge for previ-
ous reviews.13-16 In our review insufficient published
outcome data and variations in design and outcome
reporting would have led to the exclusion of relevant
trials, because the published aggregate data lacked the
necessary detail for pooling in a meta-analysis.4,84,85
Inhaled Cannabis for Chronic Neuropathy
The meta-analysis of individual patient data and the
inclusion of additional recently published RCTs increased
the power of our evidence synthesis and greatly
increased the confidence in the effect of inhaled
cannabis for chronic neuropathy compared with previ-
ous reviews.46,54,55,67 The Bayesian posterior probability
of more than 99.7% indicates the very high likelihood
that inhaled cannabis is effective in the short term for 1
in 5 or 6 patients with chronic neuropathic pain
(Supplementary Box 1), unlike the classic P value, which
indicates how unlikely the observed outcomes data are
given a null hypothesis of no effect. To our knowledge,
this is the first Bayesian meta-analysis of individual pa-
tient data in medicine.6
The Observed Short-Term Effect of Inhaled
Cannabis Is Meaningful for 1 in 5 or 6 Patients
With Chronic Neuropathy
Our responder analysis is showing a statistically signif-
icant and minimal clinically important difference for 1 in
5 to 6 patients, an effect measure easily understood by
patients, payers, and providers alike.58 Responder anal-
ysis has been advocated for patient-reported outcomes
in chronic pain trials to distinguish a minimal but statisti-
cally significant difference between groups on a popula-
tion basis from a clinically meaningful effect for the
individual participant.28,60,61 Our cutoff for a
meaningful response (>30%) is 1) grounded in what
patients themselves judge to be important
improvement32 and 2) based on expert consensus (IM-
MPACT).58 Based solely on frequentist hypothesis
testing, responder analysis may miss the goal, while
losing power.72 Our Bayesian meta-analysis of individual
patient data allowed us to calculate a posterior probabil-
ity of effect larger than 99.7%.
Limitations
Effects Are Consistent Across Different Causes
and Populations
We pooled data from populations with chronic painful
neuropathy of different causes and in different popula-
tions. We included HIV-related distal sensory polyneurop-
athy, posttraumatic, complex regional pain syndrome,
peripheral and diabetic peripheral neuropathy, and pa-
tients with and without previous exposure to cannabis.
Similar approaches were also taken by authors of previ-
ous reviews on cannabis for chronic painful neuropa-
thy.13-16 Evidence synthesis across distinct but closely
related painful neuropathies is reasonable because
their clinical course and pathological mechanism are
considered similar and receive uniform treatment
recommendations9,64; Indeed, the ‘‘etiological factors
responsible for driving the mechanisms are not disease
specific’’ and ‘‘disease diagnosis is not helpful in
selecting the optimal pain therapy’’.86,87 Even if the
absence of evidence for heterogeneity constitutes no
evidence for clinical homogeneity,44 the consistency
and uniformity of the effect of inhaled cannabis on
chronic neuropathic pain across different causes and
Andreae et al
populations, further enhances our confidence in the
generalizability of our findings.48 Yet, our meta-analysis
can only be as strong as the underlying data (Tables 1
and 2) and the methodological quality (Fig 2 and
Supplementary Table 1); the small number of studies
included, their small number of participants, and short-
comings in allocation concealment42 and attrition
(Table 2) limit our ability to draw firm conclusions. The
small numbers of studies found in each subgroup pre-
cluded a formal study of publication bias. A graphical
analysis or the test proposed by Egger et al29 should at
least include 10 studies because, with fewer studies, the
power of the tests is insufficient to distinguish chance
from real asymmetry.44 We find that the use of an active
placebo to mimic the psychotropic effects of experi-
mental treatments, although it improves blinding, does
not necessarily improve the evidence regarding effective-
ness in a pragmatic clinical setting, but it does acknowl-
edge the risk of performance bias.70 Also meta-analyses
of sparse data can be unstable38,66; however, our
evidence synthesis is based on individual patient data
from all included trials, the best available source of
evidence, short of a large RCT.44,76
Cannabis Dose and Mode of Administration
May Influence Pain Relief
Estimating bioavailable cannabis is difficult. Many fac-
tors influence the amount of THC per cigarette, particu-
larly whether the material is dry or freshly picked
(Supplementary Table 3). The dose delivered likely differs
from what is actually ingested69; we validated our dose
estimates with the primary authors of the studies
included. In the forest plot of the raw responder data, a
higher dose seems to be associated with a stronger effect
(Fig 3). Our sensitivity analysis controlling for cannabis
dose only marginally improved the precision (data not
shown); hence, at the individual patient level, the dose
differences did not explain the differences in effect. This
may effectively reflect the individual dose titration.
We cannot comment on long-term adverse effects
because the available trials followed their patients for
a maximum of 2 weeks.3 Recently, several authors have
raised concerns about driving while intoxicated, with-
drawal, addiction, adverse cardiovascular, pulmonary
and cognitive effects, especially in the developing brain,
although several of these misgivings remain conten-
tious.11,16,17,26,37,40,47,68,77,83,88 Extrapolating from
recreational use is problematic and the risk-benefit bal-
ance differs when pain is medically intractable. Clearly,
we need to learn more about the benefits and risk asso-
ciated with long-term cannabis use.
Our Bayesian Meta-analysis Is Robust to
Parameter Choices and Model Assumptions
Bayesian methods are sometimes critiqued for their
presumed subjectivity, but the short-term effects of
inhaled cannabis for about 1 in 5 patients with chronic
neuropathic pain are robust and independent of our
mode of evidence synthesis. Our assumptions are
modeled explicitly and tested.14 Priors for our meta-
The Journal of Pain 1229
analysis were uninformative in order to minimize
subjectivity and just served to ensure computational
convergence. As detailed in the results and illustrated
in Supplementary Box 2, when subjected to sensitivity
analysis, our findings were robust to the choice of pa-
rameters and models. Unsurprisingly, running a fre-
quentist analysis resulted in similar estimates, except
that the CRIs of our Bayesian estimate were more con-
servative because they were based on more cautious
between-study variance estimates. Obviously, the
Bayesian approach provides a posterior probability
(99.7%; for the short-term benefits of inhaled cannabis
for about 1 in 5 patients with chronic neuropathy), an
inference not possible in the frequentist paradigm.74
The result of any meta-analysis will critically depend
on this estimate of the between-study variance. Our
between-study variability estimation was more conser-
vative than the classic random-effects approach pro-
moted by the Cochrane Collaboration,43 which itself is
more conservative than the often employed fixed ef-
fects model. Indeed, the continued debate on fixed
versus random-effects models, concerns about assump-
tions, and underestimation of between-study vari-
ability24 demonstrate that the classic ‘‘frequentist’’
statistical approach is also not free of subjectivity.78
The use of subjective model parameters destroys the
illusion of objectivity in ‘‘frequentist’’ as well as in
meta-analysis.78 Our Bayesian approach transparently
included any subjective choice explicitly in our model
and subjected all to sensitivity analysis.22,66
Recommendations for Future Research
We lack long-term pragmatic clinical trials to deter-
mine if the effects of cannabis on chronic painful neurop-
athy are sustained and durable, if cannabis use is feasible
in the community given the associated stigma,1,18 if
cannabis can be safely prescribed in vulnerable and
young populations,26,77 and if long-term adverse effects
outweigh the benefits of inhaled cannabis.11,40,47,68,77
Although the cost of inhaled cannabis is likely to be
low, medicinal cannabis continues to be controversial
(indeed illegal in many jurisdictions) and patients may
vary in their preferences on inhaling cannabis,
especially as long as it remains stigmatized. We need to
investigate if individual titration allows for the best
balance of beneficial to adverse effects. The effects of
cannabis for other conditions should equally be
explored in publicly funded rigorous RCTs. Solid clinical
evidence may facilitate selective prescribing, prevent
misuse, and reduce opioid-related harms.41
Conclusions
Our meta-analysis of individual patient data suggests
that inhaled cannabis results in short-term benefits for
chronic neuropathic pain with an NNT of 5.6 [3.4,
13]CRI95% (Fig 3). We lack evidence regarding sustained
long-term benefits and risks in the community setting.
The small number of studies and participants included
in the analysis (Table 1) and the risk of detection and
1230 The Journal of Pain
performance bias weaken our ability to draw firm con-
clusions (Fig 2). In our responder analysis of the propor-
tion of patients with at least 30% reduction in chronic
pain as the minimal clinically important difference, a
meaningful improvement at the individual patient level
was found for about 1 in every 5 to 6 patients
treated.27,58 This effect on chronic painful neuropathy
is consistent across diverse causes, all hitherto resistant
to available treatments (Table 1). To our knowledge,
ours is the first Bayesian meta-analysis of individual pa-
tient data. The Bayesian modeling approach may be flex-
ibly extended to other fields and questions where
References
1. Abrams DI: Medical marijuana: tribulations and trials. J
Psychoactive Drugs 30:163-169, 1998
2. Abrams DI, Jay CA, Shade SB, Vizoso H, Reda H, Press S,
Kelly ME, Rowbotham MC, Petersen KL: Cannabis in painful
HIV-associated sensory neuropathy: a randomized placebo-
controlled trial. Neurology 68:515-521, 2007
3. Andreae MH, Andreae DA: Regional anaesthesia to pre-
vent chronic pain after surgery: a Cochrane systematic re-
view and meta-analysis. Br J Anaesth 111:711-720, 2013
4. Andreae MH, Carter G, Johnson M, Sacks H: Bayesian
methods improve evidence synthesis for complementary
and alternative medicine: cannabis for painful HIV-related
distal sensory polyneuropathy (HIV-DSPN). Reg Anesth
Pain Med 36:E178, 2011
5. Andreae MH, Sacks H, Indyk D, Carter G, Johnson M:
Cannabis for HIV related chronic neuropathy. Available at:
http://www.crd.york.ac.uk/PROSPERO/display_record.asp?
ID=CRD42011001182. Accessed October 1, 2015
6. Ashby D: Bayesian statistics in medicine: a 25 year review.
Stat Med 25:3589-3631, 2006
7. Ashby D, Smith AF: Evidence-based medicine as Bayesian
decision-making. Stat Med 19:3291-3305, 2000
8. Ashton JC, Milligan ED: Cannabinoids for the treatment
of neuropathic pain: clinical evidence. Curr Opin Investig
Drugs 9:65-75, 2008
9. Attal N, Cruccu G, Baron R, Haanpaa M, Hansson P,
Jensen TS, Nurmikko T: EFNS guidelines on the pharmaco-
logical treatment of neuropathic pain: 2010 revision. Eur J
Neurol 17:1113-e1188, 2010
10. Azorlosa JL, Heishman SJ, Stitzer ML, Mahaffey JM:
Marijuana smoking: effect of varying delta 9-
tetrahydrocannabinol content and number of puffs. J Phar-
macol Exp Ther 261:114-122, 1992
11. Barber PA, Roberts S, Spriggs DA, Anderson NE: Adverse
cardiovascular, cerebrovascular, and peripheral vascular ef-
fects of marijuana: what cardiologists need to know. Am J
Cardiol 113:1086, 2014
12. Baron R, Binder A, Wasner G: Neuropathic pain: diag-
nosis, pathophysiological mechanisms, and treatment. Lan-
cet Neurol 9:807-819, 2010
13. Beard S, Hunn A, Wight J: Treatments for spasticity and
pain in multiple sclerosis: a systematic review. Health Tech-
nol Assess 7:iii. ix-x, 41-111, 2003
Inhaled Cannabis for Chronic Neuropathy
variance in outcome reporting hampers the classic
approach to meta-analysis.6,7,78
Acknowledgments
We gratefully acknowledge the assistance of Maud
Dupuy and Kirti Dasu.
Supplementary Data
Supplementary data related to this article can be
found at http://dx.doi.org/10.1016/j.jpain.2015.07.009.
14. Best N, Ashby D, Dunstan F, Foreman D, McIntosh NA:
Bayesian approach to complex clinical diagnoses: a case-
study in child abuse. J R Stat Soc Ser A Stat Soc 176:53-96,
2013
15. Birse F, Derry S, Moore RA: Phenytoin for neuropathic
pain and fibromyalgia in adults. Cochrane Database Syst
Rev;CD009485, 2012
16. Bostwick JM: Blurred boundaries: the therapeutics and
politics of medical marijuana. Mayo Clin Proc 87:172-186,
2012
17. Bostwick JM, Reisfield GM, DuPont RL: Clinical decisions.
Medicinal use of marijuana. N Engl J Med 368:866-868, 2013
18. Bottorff JL, Bissell LJ, Balneaves LG, Oliffe JL,
Capler NR, Buxton J: Perceptions of cannabis as a stigma-
tized medicine: a qualitative descriptive study. Harm Re-
duct J 10:2, 2013
19. Campbell FA, Tramer MR, Carroll D, Reynolds DJ,
Moore RA, McQuay HJ: Are cannabinoids an effective and
safe treatment option in the management of pain? A qual-
itative systematic review. BMJ 323:13-16, 2001
20. Carlin JB: Meta-analysis for 2 x 2 tables: a Bayesian
approach. Stat Med 11:141-158, 1992
21. Chaparro LE, Wiffen PJ, Moore RA, Gilron I: Combination
pharmacotherapy for the treatment of neuropathic pain in
adults. Cochrane Database Syst Rev 7:CD008943, 2012
22. Cohen HW: P values: use and misuse in medical litera-
ture. Am J Hypertens 24:18-23, 2011
23. Colombo B, Annovazzi PO, Comi G: Medications for
neuropathic pain: current trends. Neurol Sci 27:S183-S189,
2006
24. Cornell JE, Mulrow CD, Localio R, Stack CB,
Meibohm AR, Guallar E, Goodman SN: Random-effects
meta-analysis of inconsistent effects: a time for change.
Ann Intern Med 160:267-270, 2014
25. Corrigan R, Derry S, Wiffen PJ, Moore RA: Clonazepam
for neuropathic pain and fibromyalgia in adults. Cochrane
Database Syst Rev;CD009486, 2012
26. DuPont RL, Lieberman JA: Young brains on drugs. Sci-
ence 344:557, 2014
27. Dworkin RH, O’Connor AB, Backonja M, Farrar JT,
Finnerup NB, Jensen TS, Kalso EA, Loeser JD, Miaskowski C,
Nurmikko TJ, Portenoy RK, Rice AS, Stacey BR, Treede RD,
Turk DC, Wallace MS: Pharmacologic management of neuro-
pathic pain: evidence-based recommendations. Pain 132:
237-251, 2007
Andreae et al
28. Dworkin RH, Turk DC, McDermott MP, Peirce-
Sandner S, Burke LB, Cowan P, Farrar JT, Hertz S,
Raja SN, Rappaport BA, Rauschkolb C, Sampaio C: Inter-
preting the clinical importance of group differences in
chronic pain clinical trials: IMMPACT recommendations.
Pain 146:238-244, 2009
29. Egger M, Davey Smith G, Schneider M, Minder C: Bias in
meta-analysis detected by a simple, graphical test. BMJ 315:
629-634, 1997
30. Ellis RJ, Rosario D, Clifford DB, McArthur JC, Simpson D,
Alexander T, Gelman BB, Vaida F, Collier A, Marra CM,
Ances B, Atkinson JH, Dworkin RH, Morgello S, Grant I:
Continued high prevalence and adverse clinical impact of
human immunodeficiency virus-associated sensory neurop-
athy in the era of combination antiretroviral therapy: the
CHARTER Study. Arch Neurol 67:552-558, 2010
31. Ellis RJ, Toperoff W, Vaida F, van den Brande G,
Gonzales J, Gouaux B, Bentley H, Atkinson JH: Smoked me-
dicinal cannabis for neuropathic pain in HIV: a randomized,
crossover clinical trial. Neuropsychopharmacology 34:
672-680, 2009
32. Farrar JT, Troxel AB, Stott C, Duncombe P, Jensen MP:
Validity, reliability, and clinical importance of change in a
0-10 numeric rating scale measure of spasticity: a post hoc
analysis of a randomized, double-blind, placebo-controlled
trial. Clin Ther 30:974-985, 2008
33. Finnerup NB, Sindrup SH, Jensen TS: The evidence for
pharmacological treatment of neuropathic pain. Pain 150:
573-581, 2010
34. Gelman A: Prior distributions for variance parameters in
hierarchical models (comment on article by Browne and
Draper). Bayesian analysis 1:515-534, 2006
35. Gelman A, Carlin JB, Stern HS, Rubin DB: Bayesian Data
Analysis, Vol. 2. London, Chapman & Hall/CRC, 2014
36. Goodman SN: Introduction to Bayesian methods I:
measuring the strength of evidence. Clin Trials 2:282-290.
discussion 301-304, 364-378, 2005
37. Greydanus DE, Hawver EK, Greydanus MM, Merrick J:
Marijuana: current concepts. Front Public Health 1:42, 2013
38. Gunnell D, Saperia J, Ashby D: Selective serotonin reup-
take inhibitors (SSRIs) and suicide in adults: meta-analysis of
drug company data from placebo controlled, randomised
controlled trials submitted to the MHRA’s safety review.
BMJ 330:385, 2005
39. Guyatt GH, Oxman AD, Kunz R, Vist GE, Falck-Ytter Y,
Schunemann HJ: What is ‘‘quality of evidence’’ and why is
it important to clinicians? BMJ 336:995-998, 2008
40. Hashibe M, Straif K, Tashkin DP, Morgenstern H,
Greenland S, Zhang ZF: Epidemiologic review of marijuana
use and cancer risk. Alcohol 35:265-275, 2005
41. Hayes MJ, Brown MS: Legalization of medical marijuana
and incidence of opioid mortality. JAMA Intern Med 174:
1673-1674, 2014
42. Hewitt C, Hahn S, Torgerson DJ, Watson J, Bland JM: Ad-
equacy and reporting of allocation concealment: review of
recent trials published in four general medical journals.
BMJ 330:1057-1058, 2005
43. Higgins JP, Green S, Cochrane Collaboration: Cochrane
Handbook for Systematic Reviews of Interventions. Chiches-
ter, Wiley-Blackwell, 2008
The Journal of Pain 1231
44. Higgins JP Green S: Cochrane Handbook for Systematic
Reviews of Interventions 5.1.0. 2011 Available at: www.
cochrane-handbook.org
45. Hill KP: Medical marijuana for treatment of chronic pain
and other medical and psychiatric problems: a clinical re-
view. JAMA 313:2474-2483, 2015
46. Iskedjian M, Bereza B, Gordon A, Piwko C, Einarson TR:
Meta-analysis of cannabis based treatments for neuropathic
and multiple sclerosis-related pain. Curr Med Res Opin 23:
17-24, 2007
47. Joshi M, Joshi A, Bartter T: Marijuana and lung diseases.
Curr Opin Pulm Med 20:173-179, 2014
48. Kish L: Representation, randomization, and realism. in
Statistical Design for Research. Hoboken, N.J, Wiley-Inter-
science, 2004, p xxii, 267
49. Kissin I: Long-term opioid treatment of chronic nonma-
lignant pain: unproven efficacy and neglected safety? J Pain
Res 6:513-529, 2013
50. Koppel BS, Brust JC, Fife T, Bronstein J, Youssof S,
Gronseth G, Gloss D: Systematic review: efficacy and safety
of medical marijuana in selected neurologic disorders:
report of the Guideline Development Subcommittee of the
American Academy of Neurology. Neurology 82:1556-1563,
2014
51. Kruschke JK: Bayesian estimation supersedes the t test. J
Exp Psychol Gen 142:573-603, 2013
52. Leger JM: Diagnosis of chronic neuropathy. J Neurol 246:
156-161, 1999
53. Lunn D, Thomas A, Best N, Spiegelhalter D: WinBUGS - a
Bayesian modelling framework: concepts, structure, and
extensibility. Stat Comput 10:325-337, 2000
54. Lutge EE, Gray A, Siegfried N: The medical use of
cannabis for reducing morbidity and mortality in patients
with HIV/AIDS. Cochrane Database Syst Rev;CD005175, 2013
55. Lynch ME, Campbell F: Cannabinoids for treatment of
chronic non-cancer pain; a systematic review of randomized
trials. Br J Clin Pharmacol 72:735-744, 2011
56. Martin-Sanchez E, Furukawa TA, Taylor J, Martin JL: Sys-
tematic review and meta-analysis of cannabis treatment for
chronic pain. Pain Med 10:1353-1368, 2009
57. McGilveray IJ: Pharmacokinetics of cannabinoids. Pain
Res Manag 10(Suppl A):15A-22A, 2005
58. McGlothlin AE, Lewis RJ: Minimal clinically important
difference: defining what really matters to patients. JAMA
312:1342-1343, 2014
59. Moher D, Liberati A, Tetzlaff J, Altman DG: Preferred re-
porting items for systematic reviews and meta-analyses: the
PRISMA statement. Int J Surg 8:336-341, 2010
60. Moore RA, Derry S, Wiffen PJ: Challenges in design and
interpretation of chronic pain trials. Br J Anaesth 111:38-45,
2013
61. Moore RA, Moore OA, Derry S, Peloso PM,
Gammaitoni AR, Wang H: Responder analysis for pain relief
and numbers needed to treat in a meta-analysis of etori-
coxib osteoarthritis trials: bridging a gap between clinical
trials and clinical practice. Ann Rheum Dis 69:374-379, 2010
62. Moore RA, Wiffen PJ, Derry S, Toelle T, Rice AS: Gaba-
pentin for chronic neuropathic pain and fibromyalgia in
1232 The Journal of Pain
adults. Cochrane Database Syst Rev. 2014, Issue 4. Art. No.:
CD007938
63. Namaka M, Leong C, Grossberndt A, Klowak M,
Turcotte D, Esfahani F, Gomori A, Intrater H: A treatment al-
gorithm for neuropathic pain: an update. Consult Pharm 24:
885-902, 2009
64. O’Connor AB, Dworkin RH: Treatment of neuropathic
pain: an overview of recent guidelines. Am J Med 122:
S22-S32, 2009
65. Okie S: Medical marijuana and the Supreme Court. N
Engl J Med 353:648-651, 2005
66. Pereira TV, Ioannidis JP: Statistically significant meta-
analyses of clinical trials have modest credibility and inflated
effects. J Clin Epidemiol 64:1060-1069, 2011
67. Phillips TJ, Cherry CL, Cox S, Marshall SJ, Rice AS: Phar-
macological treatment of painful HIV-associated sensory
neuropathy: a systematic review and meta-analysis of rand-
omised controlled trials. PLoS One 5:e14433, 2010
68. Preuss UW, Watzke AB, Zimmermann J, Wong JW,
Schmidt CO: Cannabis withdrawal severity and short-
term course among cannabis-dependent adolescent and
young adult inpatients. Drug Alcohol Depend 106:
133-141, 2010
69. Ramaekers J, Kauert G, Theunissen E: Up in smoke:
comparability of THC dosing across performance studies.
Neuropsychopharmacology 31:2800, 2006
70. Rintala DH, Holmes SA, Courtade D, Fiess RN, Tastard LV,
Loubser PG: Comparison of the effectiveness of amitripty-
line and gabapentin on chronic neuropathic pain in persons
with spinal cord injury. Arch Phys Med Rehabil 88:1547-1560,
2007
71. Simmonds MC, Higgins JP, Stewart LA, Tierney JF,
Clarke MJ, Thompson SG: Meta-analysis of individual pa-
tient data from randomized trials: a review of methods
used in practice. Clin Trials 2:209-217, 2005
72. Snapinn SM, Jiang Q: Responder analyses and the assess-
ment of a clinically relevant treatment effect. Trials 8:31,
2007
73. Song F, Clark A, Bachmann MO, Maas J: Simulation eval-
uation of statistical properties of methods for indirect and
mixed treatment comparisons. BMC Med Res Methodol 12:
138, 2012
74. Spiegelhalter DJ, Abrams KR, Myles JP: Bayesian ap-
proaches to clinical trials and health care evaluation. Chi-
chester, Wiley, 2004
75. Stewart GB, Altman DG, Askie LM, Duley L,
Simmonds MC, Stewart LA: Statistical analysis of individual
participant data meta-analyses: a comparison of methods
and recommendations for practice. PLoS One 7:e46042,
2012
Inhaled Cannabis for Chronic Neuropathy
76. Sznitman SR, Zolotov Y: Cannabis for therapeutic pur-
poses and public health and safety: a systematic and critical
review. Int J Drug Policy 26:20-29, 2014
77. Volkow ND, Baler RD, Compton WM, Weiss SR: Adverse
health effects of marijuana use. N Engl J Med 370:2219-2227,
2014
78. Wagenmakers EJ, Lee M, Lodewyckx T, Iverson G:
Bayesian versus frequentist inference. In: Hoijtink H,
Klugkist I, Boelen P (eds): Bayesian Evaluation of Informative
Hypotheses. Statistics for Social and Behavioral Sciences.
New York, Springer, 2008, pp 181-207
79. Wallace JM: Update on pharmacotherapy guidelines for
treatment of neuropathic pain. Curr Pain Headache Rep 11:
208-214, 2007
80. Ware MA, Rueda S, Singer J, Kilby D: Cannabis use by
persons living with HIV/AIDS: patterns and prevalence of
use. J Cannabis Therapeutics 3:3-15, 2003
81. Ware MA, Wang T, Shapiro S, Robinson A, Ducruet T,
Huynh T, Gamsa A, Bennett GJ, Collet JP: Smoked cannabis
for chronic neuropathic pain: a randomized controlled trial.
CMAJ 182:E694-701, 2010
82. Wiffen PJ, Derry S, Moore RA, Aldington D, Cole P,
Rice AS, Lunn MP, Hamunen K, Haanpaa M, Kalso EA: Anti-
epileptic drugs for neuropathic pain and fibromyalgia - an
overview of Cochrane reviews. Cochrane Database Syst
Rev;CD010567, 2013
83. Wilkinson ST, D’Souza DC: Problems with the medicali-
zation of marijuana. JAMA 311:2377-2378, 2014
84. Wilsey B, Marcotte T, Deutsch R, Gouaux B, Sakai S,
Donaghe H: Low-dose vaporized cannabis significantly im-
proves neuropathic pain. J Pain 14:136-148, 2013
85. Wilsey B, Marcotte T, Tsodikov A, Millman J, Bentley H,
Gouaux B, Fishman S: A randomized, placebo-controlled,
crossover trial of cannabis cigarettes in neuropathic pain. J
Pain 9:506-521, 2008
86. Woolf CJ: Dissecting out mechanisms responsible for pe-
ripheral neuropathic pain: implications for diagnosis and
therapy. Life Sci 74:2605-2610, 2004
87. Woolf CJ, Mannion RJ: Neuropathic pain: aetiology,
symptoms, mechanisms, and management. Lancet 353:
1959-1964, 1999
88. Woolridge E, Barton S, Samuel J, Osorio J, Dougherty A,
Holdcroft A: Cannabis use in HIV for pain and other medical
symptoms. J Pain Symptom Manag 29:358-367, 2005
89. Wright S: Cannabinoid-based medicines for neurological
disorders - clinical evidence. Mol Neurobiol 36:129-136, 2007
90. Zeger SL, Liang KY, Albert PS: Models for longitudinal
data: a generalized estimating equation approach. Biomet-
rics 44:1049-1060, 1988